The present investigation has been undertaken to determine, in vivo, if nitric oxide is responsible for cytokine-induced myocardial depression. On the basis of in vitro data, we believe the negative ionotropic effects of cytokines on the heart are mediated by nitric oxide. In isolated hamster cardiac papillary muscle, this negative ionotropic effect of cytokines can be blocked by N-G monomethyl-L-arginine (NMA), a nitric oxide synthase inhibitor. Because the in vitro data demonstrated that nitric oxide synthase inhibitors prevented tumor necrosis factor (TNF)-induced myocardial depression of rapid onset and reversal, we studied a low dose of recombinant human TNF challenge in canines. This TNF challenge produces significant, early, and short-lived myocardial depression (resolved by 24 h). Surprisingly, we found that NMA did not prevent the early (up to 6 h) deleterious effects of TNF on cardiac function. In fact, during this period, TNF and NMA effects on all cardiac and hemodynamic parameters were additive (i.e., NMA did not block TNF effects). However, 24 h after TNF infusion, NMA did ameliorate the effects of TNF on some parameters such as acidbase derangements and decreases in mean arterial pressure and systemic vascular resistance. These data suggest that the early phase of TNF-induced cardiac and vascular abnormalities may not be related to nitric oxide production. However, some of the later effects of TNF may be related to the production of nitric oxide. Given the finding suggestive of a beneficial effect of NMA 24 h post TNF infusion, we evaluated the effects of nitric oxide inhibition in the setting of higher doses of TNF causing longer lasting myocardial depression. Previous experiments using TNF challenges in canines suggest that this is a reasonable hypothesis, (i.e., there may be two phases of cardiac injury). In canines, there is an early (less than 8 h), dose-independent mechanism of myocardial depression and a late (greater than 24 h) dose-dependent mechanism of myocardial depression. It was hypothesized that the inhibition of nitric oxide synthesis might not be advantageous early when myocardial depression is dose dependent. We therefore studied both prophylactic treatment (pretreatment) and therapeutic treatment (post treatment) with NMA after TNF challenge, examining both early and late time points. Treatment with NMA lowered measures of nitric oxide production in both early and late time points. At early time points, NMA given either therapeutically or prophylactically did not prevent the adverse affects of TNF. However, at 24 h, after reversal of the NMA with l-arginine, the natural substrate for nitric oxide production, prophylactic NMA ameliorated the decline in cardiac function seen with TNF challenge. These data suggest a dual effect of TNF on cardiac function. The early effect appears to be nitric oxide independent, while the later effect appears to be nitric oxide dependent. NMA is being used with cytokine therapy for cancer patients to inhibit their cardiovascular toxicity. Studies are planned in acquired immune deficiency syndrome patients to do the same. These studies will help determine the advisability of this approach.